The social network: Program combats bullying

New laws and policies to address harassment and intimidation in schools are sprouting up in every state. But can laws and polices put a stop to bullying, or do students play a role?

Psychologist Elizabeth Levy Paluck has set out to demonstrate that students can change the climate of a school from one that tolerates bullying to one that promotes positive behavior. She is leading a research program in 58 New Jersey middle schools that harnesses the power of students’ social networks to change behavior and reduce bullying.

Elizabeth Paluck

Elizabeth Levy Paluck, an assistant professor of psychology and public affairs, is pursuing research in New Jersey middle schools to test whether it is possible to change social norms regarding bullying by targeting students’ social networks. (Photo by Denise Applewhite)

The project tests the premise that the best way to change social norms is to target the most socially connected people, and then allow the change to diffuse through the group, said Paluck, an assistant professor of psychology and public affairs. “People construct their ideas of acceptable behavior by observing others, especially influential individuals,” she said. “We would like to know if we can change the culture of a school by first changing the attitudes and behavior of these individuals.”

Paluck stressed that while some of these students are “popular,” others are the unofficial leaders of non-mainstream social groups organized around a common interest such as skateboarding or glee club.

Paluck is conducting the project, which is supported by the William T. Grant Foundation, the Spencer Foundation, the Canadian Institute for Advanced Research and Princeton’s Educational Research Section, with Hana Shepherd, a postdoctoral research associate in the psychology department and Princeton’s Woodrow Wilson School of Public and International Affairs.

The first step in the project, which began in fall 2012, was to identify a school’s influential students, or “social referents.” The researchers asked the entire student body to fill out a questionnaire listing the other individuals in his or her network. Privacy was preserved via use of anonymous codes. Using a mathematical algorithm, the Paluck team then constructed models of every social network in the school and combed for individuals with lots of connections within networks.

Social network

Elizabeth Levy Paluck, assistant professor of psychology and public affairs, and her team mapped the social networks of middle school students to identify influential individuals (social referents), who were then randomly assigned to either participate or not participate in an intervention program designed to reduce conflict in schools. In a pilot study, the researchers found that students (first-degree and second-degree connections) in the networks of social referents who participated in the antibullying program were less likely to view conflict as acceptable at the school, and were less likely to be disciplined for peer conflict, compared to students in networks of social referents who did not participate in the intervention program. (Original illustration courtesy of Elizabeth Levy Paluck.)

Once identified, some of these well connected individuals were randomly selected to join a group where they learned about bullying prevention and created an intervention program to spread positive social norms through efforts such as student assemblies, posters and wristbands. “It was important that the students design the program, to ensure that it meets their needs,” Paluck said.

Prior to starting the intervention program, the researchers conducted a baseline survey of attitudes toward bullying. At the end of the 2012-13 school year, the team conducted a follow-up survey of attitudes and collected information on behaviors. While the New Jersey results are not yet in, a pilot project in a school in Connecticut found improvements in attitudes about bullying and in reduced incidents of harassment and intimidation.

The much larger New Jersey study should allow the researchers to compare schools that received the intervention to those that did not. Within each school, the researchers can compare the attitudes and behaviors of students who belong to the networks of socially influential students who participated in creating antibullying interventions to students in the networks of influential students who did not participate.

“We need to know whether we changed the climate of conflict, did we set into motion a new expectation of what behaviors are desirable, did we make an overall difference?” Paluck said. “You cannot test this with only one social network, you have to compare to other networks where intervention was not done.”

–By Catherine Zandonella